Building decarbonization necessitates renewable heating and cooling solutions such as heat pumps. Magnetocaloric heat pumps (MCHP) offer environmental and efficiency advantages but face challenges when scaling up from existing active magnetic regenerator configurations. This study highlights uneven flow resistance, porosity, and refrigerant magnetocaloric effects as key obstacles to MCHP performance in parallel multi-bed setups. To address these effects, two control strategies for the fluid flow control system were investigated: measurement feedback control and model predictive control. Results show a 36.9 % heating power improvement with measurement feedback control, though with extended control convergence times. Model predictive control achieved approximately seven times faster control convergence compared to the measurement feedback control strategy, despite exhibiting minor overshooting. Utilization factor-based model predictive control increased the heating capacity by 1.6 %–30.9 % and the COP by 1.2 %–10.7 % in scenarios with uneven flow resistance and porosity, offering computational efficiency but assuming even magnetocaloric effects between regenerators. This assumption can be addressed by outlet temperature-based model predictive control, albeit at a higher computational cost using genetic algorithm. The findings emphasize the importance of advanced control methods to scaling up MCHP in renewable energy building systems.
Read full abstract